source: other-projects/hathitrust/wcsa/extracted-features-solr/trunk/vagrant-spark-hdfs-cluster/modules/hadoop/manifests/init.pp

class hadoop {

exec { "download_hadoop":
  # Download from nearby mirror, otherwise task can time-out with default setting of '300'
  # command => "wget -O /tmp/hadoop-2.7.3.tar.gz http://apache.mirror.amaze.com.au/hadoop/common/hadoop-2.7.3/hadoop-2.7.3.tar.gz",
  command => "wget -O /tmp/hadoop-2.7.3.tar.gz   http://www-us.apache.org/dist/hadoop/common/hadoop-2.7.3/hadoop-2.7.3.tar.gz
",
  timeout => 0,
  path => $path,
#  user => $user,
#  group => $group,
  unless => "ls ${install_dir} | grep hadoop-2.7.3",
  require => Package["oracle-java8-installer"]
 }

exec { "unpack_hadoop" :
  command => "tar -zxf /tmp/hadoop-2.7.3.tar.gz -C ${install_dir}",
  path => $path,
  creates => "${hadoop_home}-2.7.3",
  require => Exec["download_hadoop"]
 }

exec { "rename_hadoop" :
  command => "ln -s ${install_dir}/hadoop-2.7.3 ${install_dir}/hadoop",
  path => $path,
  creates => "${hadoop_home}",
  require => Exec["unpack_hadoop"]
 }

exec { 'chown_hadoop':
  command => "/bin/chown -R ${user}:${group} ${hadoop_home}-2.7.3",
  path => '/bin',
  user => 'root',
  require => Exec["rename_hadoop"]
 }

file { 
  "${hadoop_home}/etc/hadoop/slaves":
  content => template('hadoop/slaves'),
  mode => 644,
  owner => $user,
  group => $group,
  require => Exec["chown_hadoop"]
 }
 
file { 
  "${hadoop_home}/etc/hadoop/masters":
  content => template('hadoop/masters'),
  mode => 644,
  owner => $user,
  group => $group,
  require => Exec["chown_hadoop"]
 }

file {
  "${hadoop_home}/etc/hadoop/core-site.xml":
  content => template('hadoop/core-site.xml'),
  mode => 644,
  owner => $user,
  group => $group,
  require => Exec["chown_hadoop"]
 }
 
file {
  "${hadoop_home}/etc/hadoop/mapred-site.xml":
  content => template('hadoop/mapred-site.xml'),
  mode => 644,
  owner => $user,
  group => $group,
  require => Exec["chown_hadoop"]
 }
 
 file {
  "${hadoop_home}/etc/hadoop/hdfs-site.xml":
  content => template('hadoop/hdfs-site.xml'),
  mode => 644,
  owner => $user,
  group => $group,
  require => Exec["chown_hadoop"]
 }

 file {
  "${hadoop_home}/etc/hadoop/yarn-site.xml":
  content => template('hadoop/yarn-site.xml'),
  mode => 644,
  owner => $user,
  group => $group,
  require => Exec["chown_hadoop"]
 }

file {
  "${hadoop_home}/etc/hadoop/hadoop-env.sh":
  content => template('hadoop/hadoop-env.sh'),
  mode => 644,
  owner => $user,
  group => $group,
  require => Exec["chown_hadoop"]
 }

file {
   [ "${hadoop_home}/hadoop_store",
     "${hadoop_home}/hadoop_store/hdfs",
     "${hadoop_home}/hadoop_store/hdfs/namenode",
     "${hadoop_home}/hadoop_store/hdfs/datanode"]:
   ensure => 'directory',
   owner => "${user}",
   group => "${group}",
   mode => 755,
   require => Exec["chown_hadoop"]
  }

file {
  "/home/${user}/.bashrc-setup-hadoop":
  content => template('hadoop/setup-hadoop.bash'),
  mode => 644,
  owner => $user,
  group => $group,
  require => [  Exec["unpack_hadoop"], File["/home/${user}"] ]
 }

file_line { "setup_hadoop_home":
    ensure => present,
    path => "/home/${user}/.bashrc",
    line => ". .bashrc-setup-hadoop",
    require => [  Exec["unpack_hadoop"], File["/home/${user}"] ]    
 }



}